From: Paul
Date: Fri, 15 Dec 2000 20:03:45 +0000
Subject: Re: [TSSP] Coils at high elevation
I wrote: >> Using the new program the f1 error for H/D=1 in >> Malcolm's catagory (a) >> coil 1 measurements is +7.7%, and for Mark's 875 >> turn coil, +16.3%. >> >> These errors can no longer be attributed to >> defects in the capacitance >> determination software and must be due either to >> an inadequate >> representation of the environment, or more >> worryingly, a problem with >> the application of internal capacitance within the >> model. boris petkovic wrote: > You mean interturn capacitance effect of a winding has > such a dramatic influence at Fo? > If so ,I don't agree .At least,It couldn't be reason > for 875 turn coil. > 16% error for 1/4 wave resonant freq.is way too much > to be caused by neglecting interturn capacities here. I guess you're thinking of the direct interturn capacitance between neighbouring turns - what I call the 'Palermo' capacitance [*]. I agree this by itself is far too small to have significant effect - even its value is in doubt due to the proximity effect. My definition of internal capacitance includes that between remote turns, a much smaller capacitance but there are more of them and they are charged to a higher voltage. Their combined effect accounts for around 5-10% of the overall capacitance at normal coil elevations. At high elevations, external capacitance is reduced and the internal is much more significant. For Malcolm's h/d=1 coil, the internal capacitance accounts for over 50% of the total E-field energy storage [+]. Therefore if either the determination of, or the implementation of, internal capacitance is faulty in any way, the effect will be up to 10 times worse at high elevations. The effect of internal capacitance on the amplitude profiles is particularly interesting, so I'm keen to use high elevation measurements as a means of validating the internal capacitance modeling. [*] A.J. Palermo, 'Distributed Capacity of Single-layer Coils', Proc IRE 1934, Vol 22, p897. Contributes around 1% or less to coil capacitance, depending on how much proximity effect you allow for. [+] Mark's coil, which exhibits a higher discrepancy in f1, has a Cint contribution of only 17%, since it has h/d=6 which gives a smaller Cint to begin with. This leads me to believe that Cint is not to blame. Cheers, -- Paul Nicholson, Manchester, UK. --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.